While the origin of the term ‘bellwether’ is unusual, the use of bellwether trials today is becoming increasingly more common in mass tort litigation. By definition, a bellwether is an indicator of trends. The term originated from the practice of tying a bell around the neck of a wether (a castrated male sheep) to induce other sheep in a flock to follow the belled-wether. Judge Robert M. Parker, who presided over the epitomic bellwether case, In re Chevron U.S.A., Inc., 109 F.3d 1016, 1019 (5th Cir. 1997), explained “the ultimate success of the wether selected to wear the bell was determined by whether the flock had confidence that the wether would not lead them astray.”
One may be curious as to how a term originating from a farming practice applies to litigation. Courts utilize a bellwether approach when large numbers of plaintiffs are proceeding on the same theory or claim and there is no other feasible way for the courts to handle the enormous caseload. The approach has been used in many cases ranging from exposure to toxic chemicals like, Agent Orange and asbestos, to products liability cases involving medical drugs and devices, such as the Dalkon shield and breast implants.
First a group of plaintiffs are chosen to represent all the plaintiffs. Thus the issues settled at the initial grouping (often referred to as a “mini-trial”) should concern common claims or theories, such as causation and liability, among all the plaintiffs. These representative plaintiffs may proceed through a normal trial including discovery, pretrial motions, and so on. The results of the trial act as the bellwether for the other plaintiffs’ trials. The verdict from this grouping is extrapolated to the remaining plaintiffs’ cases. The results may be illustrative, dispositive, or preclusive.
In illustrative trials, the actual results may be utilized for valuing groups of claims in settlements; or for dispositive motions on behalf of defendants. The plaintiffs could also simply choose to continue with their own individual trial. In dispositive trials, the verdict of the bellwether trials is extrapolated to all remaining plaintiffs. Preclusion requires that defendants who are found liable are prevented from re-litigating the issues of general liability or causation; then remaining plaintiffs need only prove actual exposure, specific causation and individual damages (but all parties must know this ahead of time as to not violate due process).
Generally, bellwether trials are not suggested for use in immature class actions where new evidence or novel legal concepts are developed. However, they may be useful to determine if a class action will be possible. Their aim is not to achieve individual justice but strives for group typicality justice, as the uncertainty of the issues (of harm to plaintiff and liability of defendant) is distributed among the plaintiffs as a group when results are averaged.
In order to initiate a workable bellwether trial, one must clearly define the common claims of all plaintiffs, identify the variables that affect the claims, select groups of bellwether plaintiffs, and the judge must clearly explain if the verdicts in the bellwether cases will be illustrative or dispositive. There is no formal rule dictating how to create a bellwether group, but deciding which variables are used is of great importance.
There are many ways to divide plaintiffs into representative classes, some examples are: gender, work history, personal vs. property damage, type of defect claimed, length of exposure, known or unknown source, amount of damages claimed, severity of injury, type of injury or diagnosis, systemic diseases or local injury, date of exposure, or place of exposure.
At the advent of bellwether trials, courts typically allowed both sides to pick an equal number of plaintiffs, or allowed the plaintiff to choose the plaintiffs for the bellwether group. However, this resulted in bellwether groupings with inadequate representation of the larger sample. Courts now rely on statistics to choose a representative group using a random sample from the whole, preferably a stratified random sample.
Adding more variables to each group will typically lower the award by bringing in lower claim plaintiffs. Creating too many bellwether groups may end up with groups that are too small to be useful. In general, one could also argue about the inherent subjectivity used when the court tries to objectively choose the relevant variables. Rather, the court should employ the use of statisticians to achieve a true random sampling of the plaintiffs. Defendants should object if the reference class is chosen without a true scientific method to extrapolate the variables to the remaining plaintiff population. In order for the bellwether group to be useful they must be representative of the rest of the cases; the more representative the test cases, the more reliable the information about similar cases will be.
Applying a verdict to non-bellwether, or remaining, claims should only be allowed if there is scientific, statistical evidence that identifies the variables involved; and if the sample is a sufficient size to permit a finding that there is a sufficient level of confidence based on representativeness and the results obtained reflect results that would be obtained from trials of the whole. See In re Chevron, 109 F.3d at 1020.
To avoid non-mutual issue preclusion, defendants can argue against the first of the four required parts to preclusion: 1) the same issue must have been fairly and fully litigated in the previous preceding; 2) resolution of the issue must have been necessary to the outcome of the first action; 3) the plaintiff must not have been able to join the first action easily, and; 4) application of preclusion is not unfair to defendant. Defendants can claim that since all plaintiffs are objectively different by nature of human variance (i.e. previous income, age, and susceptibility to injury) the first part of the rule is not satisfied.
Defense objections on the lack of a jury for the extrapolated cases may not be a successful argument against bellwether trials on the federal level, as the jury right is achieved through one bellwether case and exists for the other plaintiffs by association. However, individual states may have a state due process or Seventh Amendment-type rule requiring juries for each plaintiff’s trial. Similarly, federal courts have ruled that utilizing statistical evidence to extrapolate liability and damages to other plaintiffs do not violate a defendant’s Due Process rights. See Schwab v. Phillip Morris USA, Inc.,
449 F. Supp. 2d 992, 1239–48 (E.D.N.Y. 2006).
The defense may also object to the systematic weakness in bellwether trials. For instance, the risk may be too great when creative judicial solutions place defendants “under intense pressure to settle”, which would obviate the possibility of appellate review. In re Rhoune-Poulec Rorer, 51 F.3d 1293, 1298 (7th Cir. 1995). In fact some courts have likened bellwether trials to “judicial blackmail”. Castano v. American Tobacco Co., 84 F.3d 734, 746 (5th Cir. 1996).
Despite many shortcomings, bellwether trials do provide some information. They force plaintiffs' counsel to do the work needed to prepare their standard trial package, and the early trials give some sense of how sound that package is. The bellwether trials force the court to resolve legal questions that arise only as a trial actually approaches and witnesses begin to take the stand. Also, bellwether trials test the expert witnesses and give both parties a sense of how much it costs to try a case. Certainly, bellwether trials are not the perfect way to resolve mass torts, but they are among the best practical solution that a multidistrict litigation judge has to offer when forced with the prospect of hundreds or thousands of similar claims.